Project Lazarus: Communitybased Overdose Prevention in Rural North Carolina

Pain Medicine 2011; 12: S77–S85 Wiley Periodicals, Inc. Project Lazarus: Community-Based Overdose

Prevention in Rural North Carolinapme_1128 77..85 Downloaded from https://academic.oup.com/painmedicine/article-abstract/12/suppl_2/S77/1918825 by guest on 29 October 2018

Su Albert, MD, MPH,*†‡ Fred W. Brason II, Results. Preliminary unadjusted data for Wilkes Chaplain,*‡§ Catherine K. Sanford, MSPH,* County revealed that the overdose death rate Nabarun Dasgupta, MPH,¶ Jim Graham,‡ and dropped from 46.6 per 100,000 in 2009 to 29.0 per †‡ Beth Lovette, MPH 100,000 in 2010. There was a decrease in the number of victims who received prescriptions for the sub- *Project Lazarus, Moravian Falls, North Carolina; stance implicated in their fatal overdose from a Wilkes County physician; in 2008, 82% of overdose †Wilkes County Health Department, Wilkesboro, North decedents received a prescription for an opioid Carolina; analgesic from a Wilkes prescriber compared with 10% in 2010. ‡Northwest Community Care Network, Winston-Salem, Conclusions. While the results from this North Carolina; community-based program are preliminary, the number and nature of prescription opioid overdose §Wilkes Healthy Carolinians Council, Wilkesboro, North deaths in Wilkes County changed during the inter- Carolina; vention. Further evaluation is required to understand the localized effect of the intervention and its ¶Department of Epidemiology, Gillings School of potential for replication in other areas. Global Public Health, University of North Carolina at Key Words. Overdose; Prescription Monitoring; Chapel Hill, Chapel Hill, North Carolina, USA Opioids; Chronic Pain; Community-Based Research; Surveillance Reprint requests to: Fred W. Brason II, Chaplain, Project Lazarus, P.O. Box 261, Moravian Falls, NC 28654, USA. Tel: 336-667-8100; Fax: 866-400-9915; Introduction E-mail: [email protected]. In response to some of the highest drug overdose death rates in the country, Project Lazarus developed a community-based overdose prevention program in Abstract Western North Carolina. Wilkes County is one of the largest land mass counties in North Carolina, covering Background. In response to some of the highest over 700 square miles in the foothills of the Appalachians drug overdose death rates in the country, Project with a current population of approximately 66,500. His- Lazarus developed a community-based overdose torically, logging, textiles and manufacturing, and cattle prevention program in Western North Carolina. The and chicken farming have been primary industries. In the Wilkes County unintentional poisoning mortality 1930s, prohibition brought about moonshine activity; rate was quadruple that of the state’s in 2009 and Wilkes is the birthplace of National Association for Stock due almost exclusively to prescription opioid pain Car Auto Racing, a sport with an explicit history inter- relievers, including fentanyl, hydrocodone, metha- twined with moonshine, suggesting generations of sub- done, and oxycodone. The program is ongoing. stance misuse and abuse at the margins of the law. With much physically demanding employment, Wilkes has a Methods. The overdose prevention program signiﬁcant population that suffers the physical conse- involves ﬁve components: community activation quences of work-related injuries, with a substantial burden and coalition building; monitoring and surveillance of chronic pain. The unemployment rate consistently data; prevention of overdoses; use of rescue exceeds the national average and combined with poverty medication for reversing overdoses by community and limited educational opportunities, creates a cycle of members; and evaluating project components. Prin- socioeconomic depression. cipal efforts include education of primary care providers in managing chronic pain and safe opioid The Wilkes County unintentional poisoning mortality rate prescribing, largely through the creation of a tool kit (primarily from drug overdoses) is quadruple that of North and face-to-face meetings. Carolina’s (46.6 vs 11.0 state mortality rate per 100,000

S77 Albert et al.

Table 1 Coalitions and organizations involved in community-based response to opioid overdose deaths in Wilkes County, North Carolina

Entity Description Responsibility

1 Substance Abuse Task Force, County-level partnership Raising awareness of overdose Wilkes Health Carolinians supporting coalition building problem

Council for health actions Downloaded from https://academic.oup.com/painmedicine/article-abstract/12/suppl_2/S77/1918825 by guest on 29 October 2018 2 Chronic Pain Initiative, Regional (substate) Medicaid Clinical education on pain Northwest Community Care authority, including 70 management; policy changes Network practices and 58,000 for Medicaid beneﬁciaries; patients in six counties seed funding for community-based response 3 Wilkes County Health Local health department Data review and collection; Department authority for action; meeting facilities 4 Project Lazarus Nonproﬁt organization Coordination of efforts between organizations and individuals; school-based education; community outreach; promotion of drug treatment; evaluation

population per year in 2009) and due almost exclusively to tion of overdoses; use of rescue medication for reversing prescription opioid pain relievers [1]. Top opioids impli- overdoses by community members; and evaluating cated in deaths include fentanyl, hydrocodone, metha- project components. The last four steps operate in a done, and oxycodone; heroin is rarely suspected in cyclical manner, with community advisory boards playing overdose deaths. The average age of death is in the late the central role in developing and designing each aspect 30s, and decedents have considerable comorbid health of the intervention. conditions, including respiratory, circulatory, and meta- bolic disorders. Those who are dying are county residents At the center of Project Lazarus is the understanding that who use opioids for both medical and nonmedical reasons communities are ultimately responsible for their own and exceeded their physiologic tolerance, either directly or health and that active participation from a coalition of in combination with other licit or illicit substances [2,3]. community partners is required for a successful public health campaign. The community-based organizations Decades of studies about drug misuse and overdose primarily responsible for responding to the overdose within North Carolina have contributed to a nuanced problem in Wilkes County are presented in Table 1. Com- understanding of the nature of deaths [2,4–9]. In a study munity activation describes the concrete actions required of Medicaid beneficiaries who died of a methadone over- to bring communities together to develop a health promo- dose, state health department researchers found that in tion scheme and to build long-term social capital. The the 34 days prior to death, 30.7% of decedents had Project Lazarus model is based on previous research on codes for methadone maintenance for addiction, while community activation for health promotion, which indi- 57.7% had codes for outpatient dispensing, most likely cates that the following organizations are the most impor- for chronic pain [10]. These findings were similar to an tant for successful public health campaigns: health earlier study conducted by a Project Lazarus staff department, schools, governmental agencies, hospitals, member in conjunction with the Centers for Disease primary care clinical practices, churches, and newspa- Control and Prevention [9]. A study of a pain clinic at a pers; the following organizations have also been identified North Carolina academic center revealed that 32% of as having a role in health promotion in nonurban areas: patients exhibited behaviors associated with misuse of television stations, health-related nonprofits, substance pain medications [11]. Taken together, these and other abuse treatment centers, and colleges [12]. local data form the basis for designing the Project Lazarus prevention strategies. The overdose prevention activities and their years in operation, as designed by the community coalitions, are Drawing upon successful public health campaigns in injury outlined in Table 2. These “bottom-up” interventions were prevention, Project Lazarus created a model for prevent- designed and developed by local individuals, agencies, ing prescription opioid overdose deaths that includes the and organizations that leveraged existing resources or following five components: community activation and coa- raised awareness and funds for new programs. The role of lition building; monitoring and surveillance data; preven- Project Lazarus has been to coordinate these efforts,

S78 Project Lazarus Community-Based Overdose Prevention

Table 2 Activities and timeline of community-based prevention of overdose

No. Activity Years of Operation

Community organization and activation 1 Town hall meetings 2006–present 2 Specialized task forces 2005–present

3 Community-based leadership 2005–present Downloaded from https://academic.oup.com/painmedicine/article-abstract/12/suppl_2/S77/1918825 by guest on 29 October 2018 4 Coalition building 2005–present 5 “Managing Chronic Pain” tool kit assembled 2007–2008 Prescriber education and behavior 6 One-on-one prescriber education on pain management (“academic detailing”) 2008–2010 7 Continuing medical education sessions on pain management 2008–2010 8 Licensing actions against prescribers by state medical board 2008 9 Promotion of CSRS 2007–present Supply reduction and diversion control 10 Hospital ED opioid dispensing policy modiﬁed (e.g., limits on amount dispensed 2008–present at once, required check of CSRS for hospital ED admissions) 11 Unused medication take-back events by sheriff and police departments, with 2009–present support from DEA and SBI 12 Fixed medicine disposal sites at law enforcement offices 2011 13 Hiring and training of drug diversion specialized law enforcement officers 2009–present Pain patient services and drug safety 14 Medicaid policy change: mandatory use of patient–prescriber agreements and 2010 pharmacy home 15 Support groups for pain patients 2008–2009 16 ED case manager for Medicaid beneﬁciaries with chronic pain 2008–present 17 Vetting of local pain clinics and facilitation of specialized pain clinic referrals 2008 Drug treatment and demand reduction 18 Drug detox program 2000–present 19 Negotiation and support for opening of satellite office-based drug treatment clinic 2009 (buprenorphine) Harm reduction 20 Naloxone prescription 2010 Community-based prevention education 21 School-based education, including pledge cards 2009–present 22 Red Ribbon campaign—warnings not to share attached to dispensed prescription 2010 packages 23 Billboard containing message against sharing medications 2010 24 Presentations at colleges, community forums, civic organizations, churches, etc. 2007–present 25 Radio and newspaper spots 2006–present

CSRS = Controlled Substances Reporting System; DEA = Drug Enforcement Administration; ED = emergency department; SBI = State Bureau of Investigation.

including developing strategic and action plans, training Interventions community organizers, and raising awareness of the overdose problem. This model operates in contrast to other Development of Overdose Prevention Efforts “top-down” public health approaches in which interventions are devised by expert advisory boards and health Table 2 lists interventions to prevent overdose fatalities in authorities, funded centrally, and subsequently, adopted Wilkes County. It is beyond the scope of this article to at the local level. There are advantages and limitations provide details of each intervention, but we highlight a to both approaches. In this article, we describe a handful later. Overdose prevention efforts were proposed community-based intervention in an Appalachian county by coalition members and developed and implemented by with high overdose rates and provide qualitative process responsible parties through their professional responsibili- observations and preliminary quantitative results. ties or by volunteers. Of particular interest, two of the

S79 Albert et al. central efforts of the Chronic Pain Initiative (CPI) have been examiner data, but no thorough evaluation has been con- to educate primary care physicians in managing chronic ducted to date [14]. However, an evaluation of the CPI has pain in the outpatient setting and in safely prescribing been completed by researchers at Wake Forest University opioid medications. Both were done largely through the and is being published separately [15]. creation of a physician’s tool kit for chronic pain management and face-to-face meetings with physicians. Evaluation of Program Components

Additional prevention efforts were designed by other orga- Community Activation and Coalition Building nizations in Wilkes County. For example, policy changes in Downloaded from https://academic.oup.com/painmedicine/article-abstract/12/suppl_2/S77/1918825 by guest on 29 October 2018 the hospital emergency department (ED) were imple- As a result of the heightened community awareness, acti- mented by hospital administrators. The Northwest Com- vation, and community-building activities, many organiza- munity Care Network (NCCN), the region’s Medicaid tions are now engaged in responding to the overdose authority, and hospital system jointly placed a case epidemic in Wilkes County. A central community organizer manager in the ED to coordinate care for chronic pain holds positions as part of Project Lazarus, the CPI, and patients who are on Medicaid or who are uninsured, the Substance Abuse Task Force and is responsible for including active follow-up for referrals to primary or sub- coordinating overdose prevention efforts and minimizing specialty care for treatment of the underlying cause of their duplication. Developments from one group are dissemi- chronic pain. In a pilot project by the NCCN, patient– nated to others, and major community-wide decisions are prescriber agreements were mandated for a subset of brought before each of the advisory boards. The commu- chronic pain patients on Medicaid in Wilkes County. As nity boards have ongoing engagement in the prevention part of their agreement, patients were locked into using a efforts in Wilkes as they review and evaluate results of single pharmacy and single prescriber for all opioid intervention, making adjustments to program elements therapy, and increased linkages were set up to facilitate when necessary. In this manner, the community boards communication between physicians. This system was are active in dictating the direction of change that they intended to place responsibility for prescribing opioids in would like to see and have shown sustainability beyond the hands of a single physician, who would be aware of all the initial charges when they were convened. the concomitant medications and patient history. Monitoring and Surveillance Monitoring Data Sources Preliminary unadjusted data from Wilkes County suggest Data from four state government-run health sources con- that the overdose death rate has dropped from 43 per stitute the core monitoring elements that are used to 100,000 in 2008 to 29 per 100,000 in 2010 (Figure 1). describe and characterize overdoses: ED visits for sub- While it is too early to draw a conclusion from these stance abuse and accidental poisonings, via North Caro- numbers alone, they are indicative of a response from lina’s mandatory syndromic surveillance infrastructure community-based prevention efforts; Wilkes County did known as the North Carolina Disease Event Tracking and not see the increases in overdose deaths that nearly every Epidemiologic Collection Tool (NC DETECT) [13]; other county in North Carolina experienced. outpatient-dispensed controlled substances from the Controlled Substances Reporting System (CSRS or “pre- Supporting the idea of a community-level effect, we have scription monitoring program”); fatal accidental poisonings seen a decrease in the number of fatal overdose victims from the North Carolina Ofﬁce of the Chief Medical Exam- who received prescriptions for the substance implicated in iner; and vital statistics from the North Carolina State their overdose from a Wilkes County physician. Speciﬁ- Center for Health Statistics. Access to the data was indi- cally, in 2008, 82% of decedents had received a prescrip- vidually negotiated by Project Lazarus staff and cleared tion for an implicated substance from Wilkes prescribers, through respective ethical review mechanisms. dropping to 10% in 2010.

Evaluation of Interventions Prevention of Overdose

Given the complex relationships among the forces at the Various organizations have contributed to efforts to community level that impact overdose mortality, and given prevent overdoses in Wilkes County. A few indicators are the multifaceted response, it is not feasible to ascertain presented later. However, it should be kept in mind that the individual causal impact of each intervention in isola- many other efforts have paralleled these beyond the con- tion. A rigorous evaluation is under way, which empha- ﬁnes of clinical practice. sizes assessment and measurement of potential confounders in Wilkes and surrounding counties. At the Physician education has been conducted by the medical time of preparation of this manuscript, only crude (unad- director of the County Health Department, who visited half justed) rates for Wilkes County were available. Until the physicians registered with the Drug Enforcement adequate assessment of and adjustment for potential Administration in the county, representing 70% of ofﬁce confounders is conducted, these data should be inter- practices, a strategy that others have called “academic preted with caution. A news report suggested a decline in detailing.” Starting in 2007, the CPI also began work on a overdose mortality in Wilkes County in 2010 citing medical tool kit for local primary care prescribers, making it avail-

S80 Project Lazarus Community-Based Overdose Prevention Downloaded from https://academic.oup.com/painmedicine/article-abstract/12/suppl_2/S77/1918825 by guest on 29 October 2018

Figure 1 Comparison of U.S., North Carolina, and Wilkes County overdose mortality data. able to all prescribers. The tool kit contains pain manage- best prevention efforts. Responsible public health pro- ment guidelines, opioid risk assessment tools, “universal grams should have services for the community that openly precautions” for opioid prescribing [16], a sample patient– acknowledge the difficulties in changing behavior and set prescriber agreement (“pain contract”), defensive pre- realistic time frames for effects of an intervention to take scription writing, patient education materials, and effect. screening, brief intervention, and referral to treatment (SBIRT) modules [17]. While other tool kits were nationally The Project Lazarus take-home naloxone provision model available at the time CPI started [18–20], the year-long addresses the need for rescue and works as follows. A process of developing a custom tool kit for the local Wilkes County resident sees a physician for routine context contributed significantly to community mobiliza- medical care. The physician, who has been trained by tion and allowed for the development of strong bonds Project Lazarus, identifies the patient as a naloxone prior- between organizations and individuals involved in the ity patient based on criteria for overdose risk (Table 3). The response. The promotion of the CSRS was emphasized 13 priority groups and risk factors were derived from a by providing forms and support to enable clinicians to review of the known etiology of opioid-induced respiratory register to use the system at all events where clinicians depression and clinical insight. When patients agree to attended, along with continued encouragement and participate in Project Lazarus, they watch a 20-minute follow-up contacts. With 70% of prescribers registered, DVD in the physician’s office. The video covers patient Wilkes County has by far the highest rate of utilization of responsibilities in pain management, storage, and dis- the CSRS in the state (average 20%) [21]. posal of opioid medications, recognizing and responding to an opioid overdose and options for substance abuse Data from Wilkes County indicated that more than half of treatment. Project Lazarus participants then go to a pre- overdose deaths occurred in the home setting, where arranged community pharmacy and pick up a free nalox- emergency medical care was never called because one kit. The messaging in Project Lazarus materials does bystanders did not recognize that the signs and symp- not dwell on the differences between “legitimate” and toms they were witnessing meant that their loved one had “illicit” users of opioids but rather presents straightforward taken a potentially lethal overdose. In other cases, medical information that can be used to prevent an overdose services were not activated soon enough, and in this large fatality. land area county, emergency services were unable to reach the victim in time to reverse the overdose. Even in In response to high use of episodic emergent care to treat communities with the most aggressive and innovative chronic pain in Wilkes county, revised policies for dispens- drug overdose prevention programs, not everyone hears ing narcotics in the ED at the only hospital in the county or comprehends overdose risk messages, and not every- were codified and posted prominently in the waiting area one is willing or able to abstain from using pharmaceutical in the hope of deterring drug-seeking behaviors. The new opioids for nonmedical purposes. Likewise, not everyone policies state that the CSRS must be accessed for all with inadequate pain control understands the dangers patients receiving an opioid and provides for reprogram- inherent in making their own medication adjustments, ming ED software to lower the default number of units of taking other people’s medicines, or combining their medi- opioids to be dispensed (nine vs 50 in the native software). cines with other substances that could increase the Follow-up appointments are made for the next working chance of overdose. A rescue response is necessary day for referral to a pain treatment expert or primary care because overdoses are still going to occur despite the doctor. While the volume of patients in the ED decreased

S81 Albert et al.

Table 3 Project Lazarus naloxone priority Project Lazarus prevented 18 inpatient hospitalizations groups and risk factors for opioid-induced due to opioid poisoning, a favorable cost : benefit ratio would be achieved; about half that many would have to respiratory depression have been avoided if taking loss of productivity into account. However, these estimates should be interpreted Naloxone priority groups and risk factors for carefully as they do not constitute a rigorous analysis, and opioid overdose are intended to serve as a rough estimate for other areas considering implementation. Recent medical care for opioid Downloaded from https://academic.oup.com/painmedicine/article-abstract/12/suppl_2/S77/1918825 by guest on 29 October 2018 poisoning/intoxication/overdose Lessons and Potential Future Applications Suspected or confirmed history of heroin or nonmedical of Findings opioid use Ն High-dose opioid prescription ( 100 mg/day morphine Just as the physiological and anatomical bases for pain equivalence) and addiction share common neurophysiological path- Any methadone prescription for opioid naive patient ways, the community-level response to prescription opioid Recent release from jail or prison use problems must address pain and abuse/addiction Recent release from mandatory abstinence program or simultaneously. There may be legal and policy-level justi- drug detox program fications for drawing distinctions between medical and Enrolled in methadone or buprenorphine nonmedical users of prescription opioids. However, our detox/maintenance (for addiction or pain) experience has been that in a small, rural community, Any opioid prescription and known or suspected: finely delineating two groups (“legitimate” vs “illicit” or Smoking, COPD, emphysema, asthma, sleep apnea, or similar constructs) is a time-consuming task with limited other respiratory system disease returns. It is a practice that exacerbates stigma, blocking Renal or hepatic disease those who are at the greatest risk for overdose from Alcohol use receiving prevention messages and an opioid overdose Concurrent benzodiazepine use antidote, a practice that is consistent with other medical Concurrent antidepressant prescription conditions and medication regimens, such as diabetes. Remoteness from or difficulty accessing medical care The extreme ends of the spectrum of misuse and abuse Voluntary patient request behaviors are easiest to identify, e.g., large-scale doctor shoppers or physicians prescribing for profit to terminal ED = emergency department; COPD = Chronic obstructive pain patients in hospice. Our investigations of deaths in pulmonary disease. Wilkes County have revealed that the vast majority of individuals who overdose fall into a gray zone between these outlier scenarios. The reality of living in the commu- initially because of this policy, the nature of complaints nity where you work has led us down a pragmatic path of shifted over time to more serious cases that needed emer- simultaneously blending supply reduction, demand reduc- gency attention, leading to higher reimbursement rates tion and harm reduction strategies. and improved patient satisfaction scores. Researchers at Wake Forest University have completed an Health Economics Assessment evaluation of the elements in the CPI tool kit managing chronic pain [15]. Based on the results of evaluation, the A complete health economics analysis of community- chronic pain tool kit is being revamped to highlight the based overdose prevention is beyond the scope of this tools that doctors found most useful (e.g., patient– article, but preliminary numbers are available. According prescriber agreement) and modifying those perceived to to the Agency for Healthcare Research and Quality, the be less useful. The evaluation did find that physicians’ average cost of an inpatient hospitalization for opioid poi- prescribing behaviors changed after exposure to the peer- soning in North Carolina in 2008 was $16,970; Medicaid, mediated education and after receiving the tool kit. Medicare beneficiaries, and the uninsured accounted for Encouraged by these findings, the hospital ED has 74.5% of these stays [22]. The estimated loss in produc- requested that a tool kit be created for treating pain in tivity for each poisoning is $18,704 [23]. In terms of expen- emergency settings. Patients also responded that the ditures in Wilkes County, physician education and partial prescriber–patient agreement and other explicit policies of community mobilization efforts resulted in approximately the CPI helped them understand their care better, setting $25,000 per year of salary time for two part-time employ- expectations that both physicians and patients could ees. The total operating budget of Project Lazarus was meet. approximately $220,000 over the 15-month period ending in December 2010, including purchase of naloxone, evalu- The public sometimes expects law enforcement officials to ation, conference attendance, travel, overhead expenses, single-handedly address the drug overdose problem in a and salaries for seven part-time employees. Many other community. Project Lazarus recognizes that law enforce- organizations and local businesses contributed resources ment has an important role to play in reducing diversion of in terms of staff time and in-kind donations, but exact prescription opioids but also that a community-wide monetary contributions have not been determined. Even if approach includes, but does not rely solely on, law

S82 Project Lazarus Community-Based Overdose Prevention enforcement. Project Lazarus helped local law enforce- is risk compensation, whereby individuals take greater ment departments hire and train two officers dedicated to risks because of the presence of a safety mechanism. cases involving the criminal diversion of prescription Studies evaluating the decade of naloxone distribution to drugs. These officers have empowered the law enforce- heroin users in the United Sates have not revealed con- ment community to take a more proactive role in respond- vincing evidence of risk compensation [31–35]. In cities ing to the overdose problem, including organizing with large-scale naloxone prescribing and dispensing pro- medication “take-back” events and dedicated disposal grams for heroin users, opioid overdose mortality has sites. consistently decreased after implementation, suggesting that naloxone distribution programs do not lead to Downloaded from https://academic.oup.com/painmedicine/article-abstract/12/suppl_2/S77/1918825 by guest on 29 October 2018 The response to overdose deaths in Wilkes County has increases in overdose deaths [36–38]. been multifaceted and phased in over time. Some of the intervention elements may have immediate effect, The North Carolina Medical Board raised these and whereas changes in physicians’ and patients’ behaviors other concerns during a public hearing on Project can take a long time to be realized, and the impact on the Lazarus in November 2007. After questioning and delib- mortality rate may not be apparent for even longer. In eration, the board issued the following position state- terms of evaluations, linking the mortality data with the ment: “The Board has reviewed and is encouraged by, CSRS can reveal associations that require modification of the efforts of Project Lazarus, a pilot program in Wilkes an intervention to avoid unintended consequences. It was County that is attempting to reduce the number of drug recognized that criminal prescribing and diversion may be overdoses by making the drug naloxone and an educa- contributing to the greater than expected overdose rates tional program on its use available to those persons at in Wilkes County, but the association was not always risk of suffering a drug overdose. The prevention of drug clear. For example, in December 2008, the North Carolina overdoses is consistent with the board’s statutory Medical Board suspended the license of a physician who mission to protect the people of North Carolina. The was deemed to have been prescribing controlled sub- Board therefore encourages its licensees to cooperate stances negligently after investigation by undercover law with programs such as Project Lazarus in their efforts to enforcement operations. Prior to the suspension, there make naloxone available to persons at risk of suffering was suspicion that the physician was prescribing improp- opioid drug overdose” [39]. erly; however, this physician’s prescriptions were not rou- tinely implicated in overdose deaths. It was only after the suspension that individuals under his care started dying of Summary overdoses, probably due to a disruption in opioid tolerance. Other managers of practices in the county did not The five-component strategy created by Project Lazarus is want to handle patients who had been under this physi- centered around community activation and a strong coa- cian’s care as they were perceived to be problem patients lition of partners who have an active interest in preventing at best and addicted at worst. It became evident that prescription overdose deaths [3]. It capitalizes on using supply reduction practices in the absence of demand existing data sources to provide perspectives on fatal and reduction and harm reduction could paradoxically nonfatal overdoses and serves as a mechanism to evalu- increase overdoses, a finding consistent with literature on ate interventions. The multiple levels of prevention efforts illicit drugs [24,25]. In response, Project Lazarus negoti- and community-based education are intended to reach ated and supported the opening of a satellite outpatient medical care providers as well as pain patients and non- treatment program in 2009, with more than 250 opioid- medical drug users without exacerbating stigma. dependent patients currently enrolled on buprenorphine School-based prevention education targets vulnerable treatment, a demand reduction approach documented to populations and aims to shift general patterns of sub- have reduced opioid overdose deaths and drug-related stance abuse. The provision of take-home naloxone crime in other countries [26]. acknowledges that prevention efforts can fail or take years to have effect and that overdose deaths can be prevented Hypothetical concerns with providing naloxone to patients in the community. Finally, evaluations of specific interven- and drug users have been raised. One concern with pre- tions can provide input on how to improve the services. hospital administration of naloxone is the return of respi- The overall impact is under evaluation, but initial results ratory depression. During a 5-year period in San Diego, suggest that the Project Lazarus model of enhanced and 998 out-of-hospital patients received naloxone (primarily coordinated empowerment in responding to overdoses due to heroin overdose) from emergency medical services among law enforcement, physicians, and pain patients and refused transport, against medical advice. Reviews of may be making headway in reversing Wilkes County’s medical examiner records found no instances of individu- epidemic of drug overdoses. Target communities for rep- als dying of opioid poisoning within the 12 hours following licating the Project Lazarus model include those with high- naloxone administration [27,28]. Further supporting these prescription opioid unintentional poisoning rates and observations, it is well-established in animal models that some degree of community awareness and coalition- opioid tolerance is place-dependent, and there is reason building capacity. The presence of a motivated community to believe that return of opioid depression after an initial organizer, support from the medical establishment, and administration of naloxone may be different in hospital and strong data utilization practices are key components for community settings [29,30]. Another hypothetical concern replication.

S83 Albert et al.

Acknowledgments 4 Ballesteros MF, Budnitz DS, Sanford CP, et al. Increase in deaths due to methadone in North Caro- The authors would like to thank Dana Zacharias for her lina. JAMA 2003;290:40. efforts in the Wilkes County Schools, Sheriff Dane Mastin, the members of the CPI, and the Wilkes Healthy Carolin- 5 Hudson P, Barringer M, McBay AJ. Fatal poisoning ians Task Force Substance Abuse Task Force for their with propoxyphene: Report from 100 consecutive efforts. Project Lazarus would also like to thank Dr. Doug cases. South Med J 1977;70:938–42. Easterling, Dr. Elizabeth Gamble, Jessica Richardson, and Y. Montez Lane at Wake Forest University; Glenda Adams, 6 Pierson S, Hansen R, Greene S, et al. Preventing Downloaded from https://academic.oup.com/painmedicine/article-abstract/12/suppl_2/S77/1918825 by guest on 29 October 2018 Tim Whitmire, Scott Proescholdbell, Katherine Harmon, medication errors in long-term care: Results and William Bronson, and John Womble of the North Carolina evaluation of a large scale Web-based error reporting Department of Health and Human Services; Amy Ising and system. Qual Saf Health Care 2007;16:297–302. Dr. Anna Waller at NC DETECT; Dr. Marsha Ford at the Carolinas Poison Center; James Bowman at the State 7 Hall MT, Edwards JD, Howard MO. Accidental deaths Bureau of Investigation with agents White and Billings; the due to inhalant misuse in North Carolina: 2000–2008. Wilkes County Health Department; Cathy Huie of Brame Subst Use Misuse 2010;45:1330–9. Huie Pharmacy; and Holly Price of the North Carolina Board of Pharmacy. 8 Tharp AM, Winecker RE, Winston DC. Fatal intrave- nous fentanyl abuse: Four cases involving extraction of Disclosures fentanyl from transdermal patches. Am J Forensic Med Pathol 2004;25:178–81. Project Lazarus is funded and supported in part by the NCCN, the Drug Policy Alliance, Qualla Boundary Eastern 9 Sanford K. Findings and recommendations of the task Band of the Cherokee Indians, Smoky Mountain LME, force to prevent deaths from unintentional drug over- Wilkes Healthy Carolinians Council, and The Governor’s doses in North Carolina, 2003. In: North Carolina Institute. Programs and organizations affiliated with Department of Health and Human Services, Project Lazarus coalitions have also received material or ed. N.C. Injury and Violence Prevention Branch. in-kind support for overdose prevention efforts from the Raleigh, NC: NC Department of Public Health; 2004. following commercial entities: Brame Huie Pharmacy, Available at: http://www.injuryfreenc.ncdhhs.gov/ Wilkes Motor Speedway, Crime Stoppers, Kohl’s, Chick- About/TaskForcetoPreventDrugDeaths.pdf (accessed Fil-A, Biscuitville, and Holly Mountain IGA. The National April 11, 2011). Association of Drug Diversion Investigators provided funding for local law enforcement officer training and posi- 10 Whitmire T, Adams G. Unintentional overdose deaths tions. Project Lazarus received one-time funding from in the North Carolina Medicaid population: Prevalence, Purdue Pharma LP through an unrestricted educational prescription drug use, and medical care services. grant, NED101356. Purdue made no contributions to the SCHS Studies 2010;162:1–12. Available at: http:// design, implementation, or evaluation of Project Lazarus www.schs.state.nc.us/SCHS/pdf/SCHS_162_WEB_ or the preparation of this manuscript. Between January 1, 081310.pdf (accessed April 11, 2011). 2008 and December 31, 2010, Nabarun Dasgupta has been a member of risk management advisory boards for 11 Ives TJ, Chelminski PR, Hammett-Stabler CA, et al. King, Cephalon, and Covidien and served as a consultant Predictors of opioid misuse in patients with chronic to King. Su Albert, Catherine Sanford, Fred Brason, Jim pain: A prospective cohort study. BMC Health Serv Graham, and Beth Lovette have no financial disclosures to Res 2006;6:46. make. 12 Von Korff M, Wickizer T, Maeser J, et al. Community References activation and health promotion: Identification of key 1 North Carolina State Center for Health Statistics. Sub- organizations. Am J Health Promot 1992;7:110–7. stances Idendified From T-codes Involved in Poisoning Deaths of Unintentional or Undetermined Intent North 13 Hakenewerth AM, Waller AE, Ising AI, Tintinalli JE. Carolina Residents: 2000–2009. Annually generated North Carolina Disease Event Tracking and Epidemio- report. Raleigh, NC: NC Department of Health and logic Collection Tool (NC DETECT) and the National Human Services, State Center for Health Statistics; Hospital Ambulatory Medical Care Survey (NHAMCS): 2010. Comparison of emergency department data. Acad Emerg Med 2009;16:261–9. 2 Sanford CP. An unrelenting epidemic of deaths from prescription drugs in North Carolina. NCMB 14 Hubbard J. Prescription drug grip on Wilkes eased. Forum 2008;2:4–7. Wilkes Journal-Patriot. North Wilkesboro, NC: Carter- Hubbard Publishing; 2010. 3 Dasgupta N, Brason F II, Albert S, Sanford CP. Project Lazarus: Overdose prevention and responsible pain 15 Easterling D, et al. Docket ID: FDA-2009-N-0143- management. NCMB Forum 2008;1:8–12. 1143, Comment on risk evaluation and mitigation

S84 Project Lazarus Community-Based Overdose Prevention

strategies for certain opioid drugs; notice of public 28 Vilke GM, Sloane C, Smith AM, Chan TC. Assessment meeting; reopening of comment period. Available for deaths in out-of-hospital heroin overdose patients at: http://www.regulations.gov/#!documentDetail;D= treated with naloxone who refuse transport. Acad FDA-2009-N-0143-1143.1 (accessed April 11, 2011). Emerg Med 2003;10:893–6.

16 Gourlay D, Heit H. Universal precautions: A matter 29 Falls WA, Kelsey JE. Procedures that produce of mutual trust and responsibility. Pain Med context-specific tolerance to morphine in rats also 2006;7:210–1. Author reply 12. produce context-specific withdrawal. Behav Neurosci Downloaded from https://academic.oup.com/painmedicine/article-abstract/12/suppl_2/S77/1918825 by guest on 29 October 2018 1989;103:842–9. 17 Bernstein E, Bernstein J, Feldman J, et al. An evidence based alcohol screening, brief intervention and 30 Siegel S, MacRae J. Environmental specificity of toler- referral to treatment (SBIRT) curriculum for emergency ance. Trends Neurosci 1984;7:140–3. department (ED) providers improves skills and utilization. Subst Abus 2007;28:79–92. 31 Galea S, Worthington N, Piper TM, et al. Provision of 18 Katz N. Opioid Prescribing Toolkit. Oxford and New naloxone to injection drug users as an overdose pre- York: Oxford University Press; 2009. vention strategy: Early evidence from a pilot study in New York City. Addict Behav 2006;31:907–12. 19 Ballantyne J. Massachusetts General Hospital. In: Fishman SM, Abdi S, eds. The Massachusetts 32 Maxwell S, Bigg D, Stanczykiewicz K, Carlberg- General Hospital Handbook of Pain Management, 3rd Racich S. Prescribing naloxone to actively injecting edition. Philadelphia, PA: Lippincott Williams & Wilkins; heroin users: A program to reduce heroin overdose 2006:587. deaths. J Addict Dis 2006;25:89–96.

20 Fine PG, Davies A, Fishman S. The Diagnosis and 33 Piper TM, Stancliff S, Rudenstine S, et al. Evaluation of Treatment of Breakthrough Pain. Oxford and New a naloxone distribution and administration program in York: Oxford University Press; 2008. New York City. Subst Use Misuse 2008;43:858–70. 21 Project Lazarus. Project Lazarus Annual Report 2010. Moravian Falls, NC: Project Lazarus Blog; 2010. Avail- 34 Seal KH, Thawley R, Gee L, et al. Naloxone distribu- able at: http://projectlazarus.posterous.com/project- tion and cardiopulmonary resuscitation training for lazarus-annual-report-2010 (accessed April 11, 2011). injection drug users to prevent heroin overdose death: A pilot intervention study. J Urban Health 22 Agency for Healthcare Research and Quality (AHRQ). 2005;82:303–11. Healthcare cost and utilization project. State Inpatient Databases, 2008. 2010. 35 Sporer KA, Kral AH. Prescription naloxone: A novel approach to heroin overdose prevention. Ann Emerg 23 Corso P, Finkelstein E, Miller T, Fiebelkorn I, Zaloshnja Med 2007;49:172–7. E. Incidence and lifetime costs of injuries in the United States. Inj Prev 2006;12:212–8. 36 Baltimore City Health Department. Intoxication Deaths Associated with Drugs of Abuse Or Alcohol. Baltimore, 24 Bohnert AS, Nandi A, Tracy M, et al. Policing and risk Maryland: Ofﬁce of Epidemiology and Planning; 2008. of overdose mortality in urban neighborhoods. Drug January 1995 through September 2007. Alcohol Depend 2011;113(1):62–8. 37 New York City Department of Health and Mental 25 Wood E, Tyndall MW, Spittal PM, et al. Impact of Hygiene. Illicit drug use in New York City. NYC Vital supply-side policies for control of illicit drugs in Signs 2010;9:1–4. the face of the AIDS and overdose epidemics: Investigation of a massive heroin seizure. CMAJ 2003;168:165–9. 38 Schmitz S, Kane-Willis K. Heroin Use in Illinois: A Ten-Year Multiple Indicator Analysis, 1998 to 2008. 26 Emmanuelli J, Desenclos JC. Harm reduction interven- Chicago, IL: Roosevelt University, The Illinois Consor- tions, behaviours and associated health outcomes in tium on Drug Policy; 2010. France, 1996–2003. Addiction 2005;100:1690–700. 39 North Carolina Medical Board. Position Statement: 27 Vilke GM, Buchanan J, Dunford JV, Chan TC. Are Drug Overdose Prevention. Raleigh, NC: Author; 2008. heroin overdose deaths related to patient release after Available at: http://www.ncmedboard.org/position_ prehospital treatment with naloxone? Prehosp Emerg statements/detail/drug_overdose_prevention/ Care 1999;3:183–6. (accessed April 11, 2011).

S85